This proposal describes a study aimed at evaluating the relation of retinal microvascular characteristics(e.g., retinal arteriolar narrowing, arterio- venous nicking, and retinopathy) to subclinical cardiovascular disease, clinical disease, and their risk factors, in the Multi-Ethnic Study of Atherosclerosis (MESA), involving 6,500 men and women aged 45-84 years of diverse ethnicities (Caucasians, African-Americans, Hispanics and Asian Americans). It extends new population based data that show retinal microvascular characteristics, as quantified from standardized retinal photography, are predictive of incident coronary heart disease in women and incident stroke in men and women, independent of blood pressure and other cardiovascular risk factors. In the current study, we propose adding retinal photography at the first follow-up of the MESA. We will grade these photographs for the presence and severity of retinal microvascular changes based on standardized protocols, including a quantitative, computer-based measurement of retinal arteriolar caliber. We propose to test new hypotheses that link retinal microvascular characteristics and arteriolar caliber to a wide array of subclinical cardiovascular measures (including left ventricular function defined from cardiac magnetic resonance imaging, peripheral arterial function defined from radial artery tonometry, and endothelial function defined from flow-mediated vasodilation of the brachial artery), clinical cardiovascular outcomes (including coronary heart disease, congestive cardiac failure and stroke), and their risk factors (including hypertension and diabetes). This ancillary study capitalizes on the structure and wealth of cardiovascular data in MESA, and will provide critical insights into the relative contribution of microvascular disease to the pathophysiology and natural history of subclinical and clinical cardiovascular disease, in representative populations of men and women of diverse ethnicities